AI Race: China's Moonshot and the Shifting Tech Landscape

The global AI race is heating up, with recent developments from Chinese tech startups like Moonshot sending ripples through financial markets. The Bloomberg Intelligence Podcast features insights from Mandep Singh, who discusses the implications of Moonshot's new model release, its focus on token efficiency and affordability, and how it fits into the broader AI landscape.

Moonshot's Market Impact and AI Efficiency

Singh notes that while Moonshot's model is catching attention for being "good enough at a low price," it doesn't necessarily mean it has caught up to the "frontier models" from companies like OpenAI and Anthropic. He emphasizes that the actual adoption of these models will be key, and that while "you can create news, the actual adoption will have to wait and see." This development speaks to the underlying nervousness consumers and investors have about who is leading the AI race.

Asian Markets Dictating AI Trends?

The conversation shifts to the growing influence of Asian markets on global stock movements, particularly within the tech sector. Singh highlights how events in Korea are affecting U.S. indices, creating nervousness around valuations. He points to news like Apple Inc.'s inquiries to the U.S. government for qualification of a Chinese memory name, which has had a downside impact on U.S. memory chip stocks like Micron and SK Hynix. Singh believes that while fundamentals might be strong, "news will have an impact on day-to-day volatility."

The full discussion can be found on Bloomberg Podcast's YouTube channel.

Competition and the LLM Ecosystem

Singh argues that the AI race benefits from increased competition, preventing a duopoly in the LLM space. He states, "It's never going to be good from, you know, an ecosystem perspective. They will have very high margins which is definitely not something that you know all the other companies want in the ecosystem." The demand for cheaper LLMs is driving interest in open-source models, including those from China. He adds, "You want to see the cost come down as adoption goes up. That wasn't happening here, and I think that's why you are seeing so much interest in open source LLMs right now."

Emerging AI Models and Future Competition

Beyond Chinese startups, Singh points to other players making waves. Meta's announcement of its Muse models and its move towards becoming a cloud provider selling APIs externally are noted. The discussion also touches upon SpaceX's Grok, which after merging with Curser, has released a new model. Singh reiterates that while these models offer lower token pricing, they still trail the functionality of frontier models. He predicts that as the gap closes, users will route queries to the most appropriate LLM for a specific task, acknowledging that "you don't need the frontier LLM for every query."

Apple's AI Push and Market Position

The podcast segment then transitions to Apple's strategy, with Mark Herman, Bloomberg News managing editor for global consumer tech, providing insights. He discusses Apple's recent stock performance, surpassing Nvidia in market capitalization, and attributes this to upcoming product cycles including home devices, foldable iPhones, AI wearables, and deep integration of AI into hardware. Herman believes Apple's focus on on-device AI processing is a key differentiator, allowing for lower capital expenditure compared to competitors. He anticipates Apple will "make a boatload of more money" from these new product lines.

The Foldable Future and China Strategy

Herman expresses excitement for the upcoming foldable iPhone, calling it "the biggest change to the iPhone in the products history." He acknowledges the skepticism around the necessity of foldable phones, describing them as "technology showcase" items. Regarding Apple's China strategy, Herman notes that Apple will be able to roll out its AI features in China, but he believes the first generation might not be a "needle mover." However, he anticipates the second generation and new Siri AI will be "pretty good."